Blind trust in GPS - a renewed media issue

[Tezster]

I suppose a lot of drivers simply plug in an address into their GPS and off they go without a care or thought. I'm naturally distru****l of technology, and I also like to plan things in advance as much as possible. So while I agree that GPS providers need to supply routes and instructions that are as accurate as possible, a little common sense goes a long way.

[Jake]

I agree there are some limitations with GPS technology. But from my experience it is mainly user error. Maps cannot be 100% accurate and that is where common sense should prevail.

Rather than blame the GPS units the media reports should focus on how to validate your route or destination beforehand. The units can do this (show you a map of the entire route or the individual turns).

I would be curious to know what Magellan thinks of this matter. It would be possible to recreate a scenario to see if any other factors came into play. Like accident investigations or simulations do.

Edit: Tezster stole my thunder while I was distracted.

[johnnysmoke]

Here are two examples where GPS can lead you awfully astray. The unit is set to avoid tolls. From Buffalo NY to Fort Erie Ontario, estimated time 10 hours(via Kingston). Actual time. 5 minutes.

Here's another. Set preferences to "most direct route". GPS takes you on unplowed, rural route roads, while a major 400 series highway is 1 km away.

The code writers for these GPS devices are very poor. Perhaps interns or summer students.

[das75]

I'm sure if the GPS was telling me to turn down a logging road in the middle of winter, the wife would be telling me otherwise.

[gzink]

Good examples would be user error such as... On the TransCanada there are probably thousands of duplicate addresses. For example, there would be 1450 TCH in Sorrento area BC, Sicamous area BC, Revelstoke area BC, etc. etc. Even Elections Canada had someone from our area heading to vote 60km away due to that error. Someone(data entry) input 1450 or whatever TCH on the wrong area of the map/database.
A person would think the GPS was wrong and out by 50 or 100km but in reality they put in the wrong data or not detailed enough data(granular).

[57]

Originally Posted by johnnysmoke

Here are two examples where GPS can lead you awfully astray. The unit is set to avoid tolls. From Buffalo NY to Fort Erie Ontario, estimated time 10 hours(via Kingston). Actual time. 5 minutes

That's pretty funny. I tried this and set in a destination of a Hotel in Chicago. It took me via Montreal from Toronto - 1400 km out of my way. When I removed the Toll restriction, it gave me the proper directions.

Last year I was using a hand-held GPS in Spain and I was 50' from my hotel, but it wanted to take me 1 km out of my way to get there because there was no "road", but a pedestrian mall. I was in pedestrian mode. It did this a lot in pedestrian mode, almost as though it thought of me as a vehicle sometimes, but not other times... My wife and I laughed at it a lot.

A GPS is a useful tool, however, if followed blindly it can often take you out of your way or not where you wish to go. "Uncommon sense" is still a requirement in this world. On long trips, I always check directions with Google maps to make sure it makes sense.

[flaminghomer]

There is a particular side street in Kansas City, MO that was dead-ended years ago by a main road crossing it. When going to a certain restaurant, Garmin still tells you turn right onto this side street from the main road, which would involve breaking through a guard rail and dropping about 10 feet. On top of that, there is another main road a short distance further on that you can turn right on which will take you right to the restaurant.

[JamesK]

^^^^
There's similar in the Palermo area of Oakville, at the intersection of Dundas St. and Bronte Rd. A few years ago, Bronte Rd. was shifted west. When going to an address on Bronte Rd., just south of Dundas, my GPS wanted me to make a left turn through the barrier dividing east & west bound traffic on Dundas. The proper route was a bit further west and south on the new part of Bronte Rd. to a street that connects to the old section of Bronte Rd.

On another occasion, when going to a destination in the north end of Waterloo, my GPS wanted to go along King St. through town, instead of taking the Conestoga Pkwy, which has an exit near where I wanted to go.

Also, this isn't just limited to GPS. I've seen similar with Yahoo maps. I live just off Hurontario, in Mississauaga, between two side streets, both of which connect to Hurontario. Yahoo wanted me to go out onto one of the streets and go in the opposite direction, up a (long) block to a cross street which I could take to the other side street and then go back to Hurontario. All this instead of just turning the opposite way to go to Hurontario!.

[753951]

C'mon guys. It's Darwin at work here. If you see sun at high noon position and your watch says it's 11PM, do you go to sleep? Didn't think so.

These Google map examples are not pranks. They are (were) real. I saw them myself. But they are not wrong. It tells you to swim almost 4000 miles. It's up to you if you are going to do it. I guess someone at Google was trying to be funny. But what's the alternative? Drive car on the bottom of the ocean. You asked for car directions from North America to Europe. What do you expect?

And, BTW, logging roads marked as regular roads, non-existent intersections, etc. They have nothing to do with less than capable GPS coders. It's inaccurate map data causing that, and commercial sources of such data are scarce. Why you ask? Just go and try to map all the roads in 10km radius from where you live, and you'll see why. Now imagine entire countries and continents.

There are projects on the web that have entire maps done by volunteers from data donated by more volunteers. They also contain errors. It's nature of the beast. Yet, using GPS, I was able to drive 3 weeks in Kenya on treacherous roads, breaking both rear shocks and struts, leaving half of exhaust system somewhere, puncturing tires, and I'm still alive and well. It's all about Darwin.

[hugh]

Also remember that there is probably an almost infinite number of ways to get from a home in say Mississauga to a home in Whitby.

For sake of argument, lets say there are 100 Billion ways.

If the GPS comes up with the correct route then its accuracy is well above 99.999999%

The little brain in the GPS has follow routing algorithms which always have a degree of error. If the trip requires a large number of changes then the likelihood for error is higher.

Then again ask "what is error"?

Is a seasonal road that takes you directly to your destination, then is it wrong? Maybe in the Winter but not the summer.

[Jake]

In case you guy were wondering TomTom, Magellan and Garmin receivers now get their road information from Tele Atlas. Which is owned by TomTom. Prior to that is was Navteq for the Garmin and Magellan. Google used Tele Atlas too.

It used to be that Navteq was better in North America than Tele Atlas data. In fact my street was wrong in Tele Atlas as recently as last year. I reported it and it was fixed.

Magellan switched just last year to Tele Atlas. I don't think the new database has been released yet.